hi! could you try invoking the model for each TODO one at a time?
it may be a bit more difficult to get thematic consistency between items, but you'll definitely get that one-to-one correspondence between inputs and outputs.
Topic:
Machine Learning & AI
SubTopic:
Foundation Models
Tags: